** Important: run "R&P replication 2019 experiment" and "R&P replication 2021 experiment" do files first **

* Figures A.4 and A.5

* Figure A.4
tabplot amb partyid, percent yreverse height(.7) sep(partyid) bar1(bcolor(blue*0.5)) bar2(bcolor(green*0.5))  ///
  showval ytitle("") subtitle(2021 September, pos(1)) saving(21partyidxamb.gph, replace)
graph combine 19partyidxamb.gph 21partyidxamb.gph, ycommon ///
  note("Source: The China Image Survey Project") saving("19vs21partyxamb", replace)
graph display , ysize(1.2) xsize(2) scale(1.1) margins(vsmall)


* Figure A.5
tabplot amb natl_id, percent yreverse height(.7) sep(natl_id) bar1(bcolor(green*0.5)) bar2(bcolor(red*0.5)) bar3(bcolor(blue*0.5)) ///
  showval ytitle("") subtitle(2021 September, pos(1)) saving(21natl_idxamb.gph, replace)
graph combine 19natl_idxamb.gph 21natl_idxamb.gph, ycommon ///
  note("Source: The China Image Survey Project") saving("19vs21natlidxamb", replace)
graph display , ysize(1.2) xsize(2) scale(1.1) margins(vsmall)

